Computer Vision
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Thursday 10 September
We launch SenseNova-U1.5, an 8B-MoT native unified multimodal model that understands, reasons about, and generates visual content within an encoder-free and VAE-free architecture. We strengthen its visual interface through spatially coherent patch reconstruction and scale its training with carefully curated generation and editing data, improved task formulation, structural prompt enhancement, and native resolutions of up to 4K. For post-training, we optimize specialized experts for visual aesthetics, bilingual text rendering, infographic generation, and image editing, and consolidate their capabilities through multi-expert on-policy distillation. Across extensive evaluations, SenseNova-U1.5 largely advances image fidelity, text rendering, complex composition, multi-reference editing, and interleaved generation, while improving instruction following and preserving subject identity, geometry, and unmodified regions. Despite limited exposure to structured formats in its generation data, SenseNova-U1.5 generalizes effectively to long, complex, and structured visual instructions, further proving that multimodal understanding can transfer to visual planning and creation. Together, these findings position native unified modelling as a promising path towards systems that perceive, reason and create within a fully end-to-end framework. We will open-source training code, including supervised fine-tuning, reinforcement learning, and on-policy distillation.
Functional Analysis
Upper bound properties of free and related Banach lattices via operators
It is known that the free $p$-convex Banach lattice on a Banach space $X$ can be represented as a space of functions on the unit ball of $X^*$. In this way, it gives rise to certain related (larger) lattices. To investigate such lattices, we introduce a new tool, related to operators into $X$. This tool is then used to (i) determine whether the lattices in question possess, or fail, properties involving upper bounds of upward directed sets – namely, the Fatou property, and the related property of monotonic boundedness; (ii) investigate the regularity of embeddings between spaces in question. We find connections between the Fatou-like properties of ${\textrm{FBL}}^{(p)}[X]$ and the Radon-Nikodym property of $X$. In addition, we give an example of $X \subset Y$ such that ${\textrm{FBL}}^{(p)}[X]$ is not a regular sublattice of ${\textrm{FBL}}^{(p)}[Y]$.
Read the full abstract
Strongly Correlated Electrons
Effective Ionic Valence and Local Magnetic Moment in Kagome Superconductors
In order to understand the unexpected similarity and the correlated behavior in kagome superconductor families AV$_3$Sb$_5$ (A = K, Rb, Cs) and ATi$_3$Bi$_5$ (A = Rb, Cs), we investigate the Hartree-scale local electronic structure of these systems. Our result indicates that V and Ti ions are both of 2+ valence such that the corresponding itinerant carrier densities are similar, and the difference in electron count is instead reflected in their quantum fluctuating ionic magnetic moments. However, due to the frustrated lattice geometry of these materials, such local moments are difficult to experimentally observe via standard probes. For verification, we systematically introduce nonmagnetic Sn impurities to locally relieve the geometric frustration and experimentally demonstrate the existence of well-defined local magnetic moments via magnetic susceptibility and muon spin rotation or relaxation ($\mu$SR) measurements. All experiments discover a systematic increase of magnetic susceptibility upon increasing nonmagnetic impurity level. Our discovered ionic moments suggest a paradigm shift from the existing itinerant carrier-only picture to one incorporating strong correlation from local ionic spins. The associated interatomic and local-itinerant correlations offer a solid ground for the emergence of the observed rich correlated behavior in this new family of superconducting materials.
Read the full abstract
Quantum Physics
Quantifying Symmetry Breaking with Metric Adjusted Quantum Geometric Tensors
Quantifying properties of quantum states through the limits of their manipulation is a central goal of quantum resource theories. For symmetry breaking, the quantum geometric tensor (QGT) governs asymptotic pure-state conversion, but a complete characterization for general mixed states has remained elusive. Here we fully resolve this problem for finite-dimensional systems under compact Lie group symmetries in the i.i.d. asymptotic regime. Specifically, we establish a single-letter formula for the optimal conversion rate between arbitrary states, with vanishing trace-distance error, in the resource theory of asymmetry. The rate is determined by a one-parameter family of metric adjusted QGTs, obtained by rescaling quantum Fisher information (QFI) matrices that interpolate between the symmetric logarithmic derivative and right logarithmic derivative (RLD) QFIs. For pure states, the RLD endpoint recovers the standard QGT. No state-independent finite subset of this family suffices in general, even for $U(1)$ symmetry, revealing a qualitative distinction from pure-state conversion. Our formula further yields an exact pure-state distillation-rate formula in terms of a single asymmetry measure, characterizes asymptotically reversible interconversion, and identifies bound asymmetry for quantum clocks. Complementarity among members of the metric adjusted QGT family also reveals an activation mechanism: a state with zero conversion rate to a mixed clock state can still enhance another input's yield under joint processing. Our proof relies on two developments of independent interest. First, we extend quantum local asymptotic normality to unitary models with arbitrary rank and spectral degeneracy. Second, we characterize convertibility between quantum Gaussian shift models in terms of the same one-parameter family of QFIs.
Read the full abstract
Earth and Planetary Astrophysics
The Mysterious Inspiral of WASP-12b: Why Obliquity Tides Cannot Drive Orbital Decay
WASP-12b's orbit is decaying, for unknown reasons. The planet's period is shrinking more rapidly than can be attributed to equilibrium tides or dynamical tides in a main-sequence star. Planetary obliquity tides could be sufficiently dissipative to drive WASP-12b's inspiral, but would also damp the planet's obliquity, halting the decay. Millholland & Laughlin proposed that a nearby, low-mass planet ($\sim 10$ M$_\oplus$) is maintaining a large obliquity for WASP-12b, sustaining the dissipation. We re-evaluated this hypothesis, finding that the companion must be more massive than originally proposed ($\gtrsim 65$ M$_\oplus$) to absorb WASP-12b's orbital angular momentum. Radial velocity data allowed us to rule out a companion of this type. Any companions within $3$ AU have $K \lesssim 14$ m/s at $95$% confidence.
Read the full abstract
Algebraic Topology
Cusp restrictions and Bunke–Naumann invariants with level structure
Restriction to the full cusp divisor of a modular curve defines a secondary invariant whose rational indeterminacy comes from a single global modular form. For every integral weakly holomorphic level-one modular form $h$ of weight divisible by four, we prove that the imported value $[h/2]$ vanishes at every nontrivial $\Gamma_0(N)$ level. The construction localizes coefficients before completion and rationalizes only after passing to homotopy groups. At odd prime level, we identify the full cusp spectrum as a product of two real Tate factors and construct a single holomorphic weight-two correction. This correction yields an actual integral global homotopy class and a rational global source class satisfying the equality required for joint annihilation. The equality transports to every odd composite level, while even levels follow by inverting two. At level three, the Mahowald–Rezk homotopy calculation leaves only the periodic $\nu$ family in stems $8K+3$. A cusp-residue homomorphism detects its secondary value, which has exact order two throughout the periodic family. As an application, the Bunke–Naumann secondary invariants of products in bidegrees $(8k+1,8k'+2)$ vanish after passage to every nontrivial level.
Read the full abstract
Distributed Computing
GPU-CFR: 80x Faster Counterfactual Regret Minimization by Compiling the Game to Static Dataflow and CUDA Graph Replay
Counterfactual regret minimization (CFR) is one of the few large numerical workloads that still runs faster on CPUs than on GPUs. Each iteration sweeps a game tree with up to billions of states in millions of small, interdependent gather and scatter steps issued through a generic tree interface. On a GPU every kernel finishes in microseconds, so kernel launches and framework dispatch dominate the run time, and prior GPU implementations have lost to optimized CPU code. We observe that for a fixed game, everything about a CFR iteration except the numerical values is known before the first iteration runs. We propose GPU-CFR, a compiler and runtime built on this observation. It compiles any game once into static dataflow: flat edge and information-set arrays, precomputed indices, and depth-level batched passes fix the entire operation sequence, and only solver state changes between iterations. Static chance folding, depth-level execution blocks, and a dual-lane reach buffer cut the number of framework operations by up to 18.1x. Because shapes, indices, and buffer addresses never change, CUDA Graph Replay records the iteration once and replays it with a single graph launch. On one A100, across an eight-game suite that spans card games, dice games, and board games, GPU-CFR runs 29.8–80.4x faster than the fastest prior GPU CFR on the same accelerator, and 14–258x faster than LiteEFG, one of the fastest open-source CPU implementations, on the four largest games. The compiled representation carries most of that margin: on eight CPU threads with no accelerator it is already 2.2–51.1x faster than the GPU baseline. On the CPU the optimized path reproduces the reference iterates bitwise, and tree construction and graph capture pay for themselves within the first solve. GPU-CFR beats every CPU and GPU baseline on the mid-to-large games of the suite without changing the update rule.
Read the full abstract
Quantum Physics
A Chip-scale Space-time Multiplexed Gaussian Boson Sampling Processor Beyond 10,000 Photons
Gaussian boson sampling (GBS) has emerged as a leading photonic paradigmfor demonstrating quantum computational advantage. Nevertheless, state-ofthe-art GBS setups face practical barriers including stringent optical alignment, phase instability, and limited programmability, which impede scalable engineering deployment. The chip-scale space-time multiplexed architecturepromises to resolve these constraints, yet it strongly demands wafer-scale chipcapabilities to simultaneously satisfy stringent requirements on low loss, highprecision and high-speed modulation. Here we report the first chip-scale spacetime multiplexed GBS system, monolithically integrating high-speed electrooptic modulators, on-chip delay lines, and a time-space multiplexed interferometric network on a thin-film lithium niobate chip, operating at a 4-GHz clockrate with detection events of up to 11,059 photons within 1 millisecond. Beyond benchmarking quantum advantage, we further reconfigure the photonichardware into a GBS-powered world model for modelling physical dynamics,which achieves lower prediction error with fewer trainable readout parameters compared with a classical echo state network (ESN) baseline. Our resultsvalidate the feasibility of our endeavor towards scalable photonic quantumhardware, and pave the way for the versatile programmable applications offuture GBS quantum systems.
Read the full abstract
Solar and Stellar Astrophysics
Teleparallel torsion and white dwarf structure in \(f(T)=T+\Xi T^2\) gravity
We investigate how quadratic torsion modifies the equilibrium structure of white dwarfs in covariant $f(T)=T+\Xi T^2$ gravity. Static spherical configurations are calculated with a fixed equation of state for cold carbon matter, including relativistic electron degeneracy and Coulomb lattice corrections. The stellar interior is matched to a vacuum exterior, and the mass is determined from the asymptotic geometry. At fixed central density, negative couplings produce more massive and more compact configurations than general relativity, whereas positive couplings give smaller masses and larger radii. The deviations increase with central density and are more pronounced in mass than in radius. The first limiting feature of each sequence depends on the coupling. For $\Xi=-10^{18}\,\mathrm{cm}^{2}$, the sequence remains monotonic up to the electron capture density and reaches $1.508\,M_\odot$. General relativity and $\Xi=+10^{18}\,\mathrm{cm}^{2}$ instead reach mass turning points at $1.385\,M_\odot$ and $1.366\,M_\odot$, respectively. The positive extreme can be continued as an equilibrium solution to the capture density, where its mass decreases to $1.24\,M_\odot$, but this configuration lies beyond the turning point and is not the maximum mass along that sequence. These results identify a density dependent structural response to torsion that changes the stellar mass scale without modifying the matter equation of state.
Read the full abstract
Robotics
EVPeriscope: Extended Perception across Aerial and Ground Vehicles with Event-based Propeller Tracking
Reliable relative localization between aerial and ground robots is a key requirement for tightly coordinated heterogeneous teams. This can be difficult to do using conventional frame-based cameras and fiducial markers because they are sensitive to motion blur, lighting variations, and payload constraints. This paper presents EVPeriscope, an event-based perception system that enables detection, localization and control of a quadrotor using an upward-facing event camera on a ground robot by detecting the high-frequency visual signature of its propellers. This system allows the quadrotor to function as an extended perception system for the ground robot when onboard sensors exhibit degradation or occlusion. We demonstrate the capabilities of this marsupial ground-aerial system via experiments in challenging field conditions with wind speeds of up to 15 mph, in both daylight and at night. We show that the system supports localization and closed-loop navigation through dense foliage where the ground robot's sensors are occluded. Our control system for the quadrotor operates at 200 Hz entirely with onboard sensing and computation. More details and experiment videos can be found on the project page: https://ongdexter.github.io/evperiscope.
Read the full abstract
Logic
Unfriendly partitions of locally finite Borel graphs
We answer in the negative the question of Thomas, recorded by Conley, Conley–Marks–Unger, and Conley–Tamuz, of whether every locally finite Borel graph admits a Borel unfriendly partition. Our counterexample has unbounded degree and is closed on a zero-dimensional Polish space; its connectedness relation is hyperfinite, and its components are bipartite and one-ended. Every unfriendly colouring is proper. Together with a parity obstruction, this rigidity rules out Baire measurable colourings that are unfriendly on a comeager set, and measurable colourings that are unfriendly almost everywhere for a quasi-invariant probability of finite average degree. In the positive direction, a Borel graph of maximum degree at most four admits a Borel unfriendly colouring whenever each component contains a cycle or a vertex of degree at most two. This reduces the Borel problem in maximum degree three to cubic forests and, with a theorem of Conley–Marks–Unger, gives Baire measurable unfriendly colourings for all Borel graphs of maximum degree at most four.
Read the full abstract
Machine Learning
General Quantification of Covariate and Concept Shifts
Generalization under distribution shift remains a core challenge in modern machine learning, yet existing learning bound theory is limited to narrow, idealized settings and is non-estimable from samples. In this paper, we bridge the gap between theory and practical applications. We first show that existing definition of concept shift breaks when the source and target supports mismatch. Leveraging entropic optimal transport, we propose a key notion: $\gamma^{*}\!$-concept shifts, and derive a general error bound unifying covariate and $\gamma^{*}\!$-concept shifts, which applies to broad loss functions, label spaces, and stochastic labeling. We further develop estimators for these shifts with concentration guarantees, and the DataShifts algorithm, which can quantify distribution shifts and estimate the error bound in most applications - a rigorous and general tool for analyzing learning error under distribution shift.
Read the full abstract
Machine Learning
Data Scarcity and Model Sparsity: Mixtures-of-Experts Overfit More to Repeated Data
As the supply of human-written text is exhausted, it has become standard practice to repeat language model training data. Prior work has studied data repetition for densely activated Transformers, but the effects of data repetition remains largely unexplored for recently dominant sparse architectures such as Mixture-of-Experts (MoE), despite their increased compute efficiency. We vary data repetition rates across single- and multi-domain data mixes, and across MoE settings, including expert count and granularity. We consistently find, for models ranging from 80M to 1B active (8.5B total) parameters, that MoEs degrade more rapidly under data repetition. This effect increases with sparsity, dictated by total rather than active parameters. While 80M dense models can repeat data over 8x with minimal degradation, MoEs instead begin to suffer at 4x, and deteriorate rapidly, ceding their performance benefits in all-unique data settings to underperform dense models after 32x. We experiment with existing regularization methods as a potential remedy. We find that some methods, such as dropout, can mitigate overfitting. In particular, with strong masking-based regularization, MoEs are able to outperform dense models even when data is repeated more than 64 times. However, no method fully matches the performance of all-unique training data. Finally, we analyze internal mechanisms correlated with MoE overfitting in high repetition regimes, and find that MoE routing universally stabilizes early in training, and that expert specialization correlates with overfitting to repeated data. In sum, our work addresses the adverse interactions between sparsity and data repetition: we present evidence for the core mechanisms of overfitting and its potential remediation, and suggest promising avenues for future methods to reduce over-specialization in model parameters by disrupting memorization patterns.
Read the full abstract
Artificial Intelligence
Can Edge-Deployable Vision-Language Models Identify Species?
Camera traps often run in the field on edge hardware with limited or no connectivity, making small, locally-deployable vision-language models (VLMs) – not frontier-scale ones – the practically relevant class to evaluate for species identification. We test whether models in this deployment-relevant 2–8B range carry genuine taxonomic knowledge, evaluating four such VLMs (Qwen3-VL 2B/4B/8B, Gemma3 4B) against the domain-specific specialist BioCLIP (300M parameters) on a 96-species task, comparing clean iNaturalist photographs against camera-trap imagery from 6 LILA.science collections, on two independently-sampled evaluation sets. All models identify species far above chance, but every model – general-purpose or specialist – degrades sharply on field imagery (domain gaps of 9.6–26.6 percentage points, consistent across taxonomic levels and both evaluation sets), indicating the degradation reflects general image legibility rather than fine-grained discrimination failure. BioCLIP substantially outperforms every VLM tested (by 33.2–59.2 percentage points across an expanded 200-image sample for every model) despite its far smaller size, suggesting the gap reflects specialized training data rather than model scale; yet BioCLIP's own domain gap (18.0 points) is statistically indistinguishable from the best VLM's (22.3 points), suggesting the clean-to-field degradation itself is a property of the image-quality shift rather than a general-purpose-model weakness. Under open-set prompting, 5.9–9.6% of responses are syntactically valid but taxonomically nonexistent species names; the relative fabrication-rate ranking across models replicates exactly across both evaluation sets, a more robust finding than any single point estimate.
Read the full abstract
Machine Learning (Statistics)
Generative Marketing Mix Modeling: A Causal Inference Framework Linking GEO and GEM to Business Impact
Generative artificial intelligence changes how firms reach customers, but standard marketing data do not record how often users see and notice a firm's name in generated answers. We develop Generative Marketing Mix Modeling (GMMM) to estimate the causal effects of Generative Engine Optimization (GEO) and Generative Engine Marketing (GEM). For GEO, GMMM combines repeated generated answers with question counts, shares of use across generative systems, and notice probabilities. For GEM, it combines records of sponsored placements with notice probabilities. GMMM compares expected business responses under alternative treatment sequences and establishes sufficient conditions for identifying the resulting effects. We investigate the empirical performance of the proposed method using simulated answers to product recommendation in English and Japanese.
Read the full abstract
High Energy Physics (Phenomenology)
Anomalous Dimensions, Matching, and Phenomenology of Dirac Fermionic Dark Matter Effective Interactions
We explore a fermionic dark matter (DM) extension of the Standard Model Effective Field Theory (SMEFT) and establish a complete renormalisation-group framework for its phenomenological investigation. We derive the anomalous-dimension matrix of all relevant dimension-five and dimension-six operators involving Standard Model (SM) and DM fields, enabling the consistent evolution of the associated Wilson coefficients (WCs) across energy scales. By combining renormalisation-group running with matching at the relevant thresholds, we construct a robust bridge between high-scale new physics and experimental observables. We then perform a comprehensive phenomenological analysis, evaluating the contributions of these operators to observables spanning a wide range of energies and deriving constraints on the WCs from current data. We obtain stringent and complementary bounds on the DM effective field theory (DMEFT) WCs from electroweak precision observables, flavour processes, lepton-flavour-violating decays, top-quark flavour-changing neutral-current decays, and invisible meson decays. Interpreted in terms of the effective scale of new physics, the resulting limits demonstrate that current precision measurements probe energy scales ranging from the TeV regime to several tens or even hundreds of TeV, highlighting the remarkable sensitivity of indirect searches to dark-sector interactions.
Read the full abstract
Computation and Language
Distance generalization in transformers: why bother with positional encoding?
Out-of-distribution length generalization, namely to extrapolate a task from short to longer context, has been studied intensively for transformers. Here we focus on distance generalization, which probes performance when inter-token distances are changed between training and inference, while keeping a fixed context length. We construct two synthetic delay copy tasks, both involving finite distances between source and recall, where tokens are copied either fully or selectively, and test models on delays unseen during training. We address three questions: (A) Do positional encoding schemes such as RoPE and ALiBi improve distance resolution relative to no positional encoding (NoPE)? (B) How does data diversity, the number of inter-token distances seen in training, affect performance? (C) When is distance transfer learning positive or negative? We present a thorough investigation, finding that it is paramount to improve our understanding of the underlying mechanisms.
Read the full abstract
Game Theory
Existence of the Core in Approval-Based Committee Elections
We resolve the central open question in the theory of approval-based committee elections: we prove that the core is always non-empty. This means that there always exists a committee for which no coalition of voters can deviate to an alternative committee of proportional size that every member strictly prefers. This property was introduced by Aziz et al. (2017) to capture proportional representation and coalitional stability, but no known algorithm achieved it except up to a constant approximation factor or in special cases. Our approach begins by strengthening the core to core+, which requires the committee to be robust even to fractional deviations. Using LP duality, this property can equivalently be defined in terms of a payment-based certificate under which no losing candidate gathers too much support. We then define a voting rule that chooses a committee admitting a certificate that maximizes a novel harmonic entropy objective function, which rewards balanced payment distributions. We show that the payment system associated with an optimal committee certifies that the committee satisfies core+, establishing existence. The same is true for local optima, which allows us to design a polynomial-time algorithm computing a core+ committee.
Read the full abstract
Artificial Intelligence
Artificial Id: Drive and Persistent Alignment in Agentic AI
Agentic AI is moving from bounded task execution toward systems that retain consequential state, continue operating and adapt across task boundaries. That shift creates a control problem that current harnesses largely solve by hand: objectives, retries, verification, stopping rules and other behavioral transitions are specified externally. We propose an artificial id, an adaptive internal drive for determining whether behavior should continue, stop or change. In a minimal virtual Petri-dish experiment, a controller too small to perform general-purpose reasoning and receiving no task-specific behavioral objective develops useful control through differential persistence. The same mechanism selects an unintended physical strategy when that behavior persists better and later replaces a learned sensor mapping when its environmental meaning changes. These results show that adaptive direction can emerge without being explicitly specified as a behavioral objective. The same persistence that makes such adaptive agency useful can also allow misalignment, corrupted state and unintended behavior to persist across task boundaries. A scalable artificial id would carry consequential state and adaptive drive across those boundaries, making alignment a property of the continuing agentic system rather than of a model response or single trajectory. Such systems require a persistent alignment boundary over trusted observations, consequence channels, persistent state, authority, identity, provenance and hard constraints.
Read the full abstract
Machine Learning
From Protocols to Evidence: Bounded Claims for AI in Service of the Common Good
Claims that AI systems improve decisions, broaden access, reduce harm, or empower users can exceed what their evaluation supports. This gap motivates our ongoing work on RISE AI. In this paper, we examine the proposed framework in light of Pope Leo XIVs Magnifica Humanitas, particularly its treatment of human dignity, ownership, and technological power. We develop three linked propositions. First, AI adoption can reveal unmet human needs and institutional weaknesses, while deployment can change how institutions respond to them. We propose a rupture test to examine whether deployment repairs or compounds existing failures, substitutes for human support, or conceals problems behind improved performance. The test would compare deployment with existing practice and feasible human or non AI alternatives. Second, responsible AI must examine the institutional conditions surrounding a system, including who owns the infrastructure, controls its use, and bears responsibility for its consequences. Third, favorable evaluation cannot by itself establish respect for human dignity or moral and political legitimacy. We distinguish evidence-bounded deployment, which limits claims to what the evidence supports under specified conditions, from measurement-bounded governance, which identifies constraints that favorable results cannot override. We outline a proposed RISE AI framework for applying these propositions across Responsibility, Inclusivity, Safety, and Empowerment.
Read the full abstract
Analysis of PDEs
Global $W^{2,p}$ and $C^{2,\alpha}$ regularity for the sigma-$2$ equation
In this paper, we prove that continuous $2$-convex solutions of the $\sigma_{2}$ equation with positive continuous density on bounded strictly mean-convex $C^{3}$ domains with $C^{3}$ boundary values belong to $W^{2,p}$ for every $1<p<\infty$ in every dimension. If, in addition, the density is $C^{\alpha}$, $0<\alpha<1$, the solutions belong to $C^{2,\alpha}$. Moreover, two counterexamples show that neither the $C^{3}$ assumption on the domain nor that on the boundary values can in general be weakened to $C^{2,1}$.
Read the full abstract
General Relativity and Quantum Cosmology
Theoretical Aspects of Direct Waves in Kerr Black Holes: Pole-Splitting Method for Ringdown Analysis
We develop a theoretical description of direct waves (DWs) from a particle plunging into a Kerr black hole in the extreme-mass-ratio regime. A DW is characterized by a source-driven complex frequency $\omega_{\rm G}$ that reflects the particle's orbital motion, including motion inside the ergoregion. Hence, it is expected to be a probe of rich structures of Kerr black holes. We show that the contribution of the quasinormal-mode (QNM) pole sector to the DW is part of dynamical QNM excitation. We also show that the apparent horizon-mode poles cancel at every order of a consistent near-horizon expansion of the trajectory, homogeneous radial solution, and complete point-particle source. This result assumes that the point particle follows a geodesic near the horizon and leaves open the possibility of a transient DW while the particle passes near the ergosurface or the vicinity of the potential barrier. To investigate this possibility in simulated plunge waveforms, we develop a pole-splitting method that decomposes the transfer function into QNM-pole and non-pole parts. The resulting waveform decomposition avoids the frequency-dependent time shift (group delay) introduced by the QNM filtering. For quasi-circular plunges into moderately and rapidly rotating Kerr black holes, the frequency and decay rate of the non-pole waveform show qualitative agreement with our theoretical prediction over finite time intervals. These results support the potential use of DWs as probes of the ergoregion and gravitational redshift effect around black holes.
Read the full abstract
Cosmology and Nongalactic Astrophysics
A Multi-Axion Ladder Across Cosmic History: From Inflation, BBN, and Early Dark Energy to Late-Time Accelerated Expansion
What if dark energy is recurrent throughout cosmic history? In this picture, episodes of scalar-field dark energy become less surprising and more natural features of cosmic evolution. We construct a homogeneous multi-axion cosmology with a transient contribution during Big Bang nucleosynthesis, two early dark energy components before recombination, and a thawing field that supplies the present dark-energy density. As Hubble friction weakens, the fields begin to roll at successive epochs set by their potential-curvature scales. Their initial displacements affect the rolling delays and peak energy fractions. The transient fields have third-power cosine potentials, whose sextic minima allow faster-than-radiation dilution during rapid, small-amplitude oscillations. All four fields are evolved in a common Friedmann background, with their first roll and subsequent dynamics resolved numerically. With reference matter and radiation densities taken from Planck 2018, the benchmark has a nucleosynthesis-era peak fraction of approximately $0.99\%$ near $z=10^9$. The two early dark energy fields peak near $z=7.9\times10^3$ and $2.4\times10^3$, with individual fractions of $8.6\%$ and $7.8\%$; their combined fraction reaches $9.7\%$. The late-time field is normalized to supply a present fraction of approximately $0.685$, while the three transients leave a combined fraction of approximately $6.0\times10^{-7}$. An aligned two-axion example illustrates the enhanced field range available for an inflationary extension. We also outline how searches for transient contributions to the expansion rate at other epochs could constrain additional axion scales.
Read the full abstract
Hardware Architecture
AccelForge: Comprehensive Modeling and Co-Design Framework for AI Accelerators
Tensor algebra workloads, of which deep neural networks are prominent examples, are energy-intensive workloads in modern datacenter and edge deployments, making accelerators necessary to achieve energy efficiency and high throughput. To quickly evaluate and iterate on accelerator designs, we need an accelerator modeling framework that captures salient attributes of devices, circuits, architectures, workloads, as well as optimizing the mapping of the workload onto the hardware. In this paper, we introduce AccelForge, which improves upon existing accelerator modeling frameworks in capabilities, speed, and ease-of-use. AccelForge unifies and multiple works into one framework, and it includes (1) composable user-defined and user-modifiable models of devices, circuits, and architectures, (2) fast mappers that enable accurate evaluation in orders of magnitude less (computer and human) time, and (3) easy-to-use and easy-to-extend, yet still high performance, Python implementations of both the model and mapper to enable rapid research and extension to novel optimizations.
Read the full abstract
Portfolio Management
Entropic Value-at-Risk parity for tempered stable returns
We develop Entropic Value-at-Risk (EVaR) parity for tempered stable returns. EVaR-based inverse risk parity (IRP) and equal risk contribution (ERC) portfolios are constructed using multivariate normal tempered stable models and independent component analysis with tempered stable components. We derive the corresponding asset-level EVaR and EVaR-deviation contributions and use the latter to separate the fitted location term from EVaR risk contributions. Under Gaussian returns, EVaR-deviation IRP and ERC recover conventional volatility IRP and ERC weights. We evaluate the resulting portfolios in three investment universes. Empirically, EVaR-based ERC portfolios achieve positive Sharpe differences relative to equal weight across the universes.
Read the full abstract
Machine Learning
TART: A Modular Tool for Technique-Aware Audio-to-Tablature Guitar Transcription
Automatic Music Transcription (AMT) for guitar remains limited by three challenges: existing systems often fail to capture expressive techniques such as slides, bends, and percussive hits; they often assign notes to incorrect string-fret combinations; and they are typically trained on clean recordings, limiting their generalization to noisy real-world audio. To address these challenges, we propose TART, a modular four-stage audio-to-tablature pipeline consisting of (1) an audio-to-MIDI transcription model, (2) an expressive technique classifier, (3) an audio-conditioned T5 encoder-decoder for string-fret assignment, and (4) an automated tablature generator. We evaluate TART in a zero-shot setting on GuitarSet, EGDB, and two augmented benchmarks, Noisy GuitarSet and Noisy EGDB. Averaged across these four benchmarks, TART achieves 81.35% audio-to-MIDI F50 (+6.67 points over the best prior baseline), 71.8% string-fret Tab F1 (+8.5 points over the best prior baseline), and 54.08% end-to-end Tab F1. To our knowledge, TART is the first framework to generate guitar tablature with both fingering and expressive technique annotations directly from guitar audio.
Read the full abstract
Quantum Physics
The generalised semi-Clifford conjecture is false
The Clifford hierarchy is a nested sequence of sets of quantum gates that can be fault-tolerantly performed using gate teleportation within standard quantum error correction schemes. The importance of these gates has motivated numerous studies of their structure. Zeng-Chen-Chuang conjectured in 2007 that all hierarchy gates are generalised semi-Clifford, i.e. take the form $C_1 \Pi D C_2$ for Clifford gates $C_1, C_2$, a permutation gate $\Pi$, and a diagonal gate $D$; Beigi-Shor proved in 2008 that this holds for all third-level gates. We construct a five-qubit gate that is in the fifth level of the Clifford hierarchy but is not generalised semi-Clifford. Rather than simply present and verify our counterexample to the generalised semi-Clifford conjecture, we show how its form can be deduced. Our counterexample also demonstrates that the Clifford hierarchy is not closed under inverses.
Read the full abstract
General Relativity and Quantum Cosmology
Energy accreted onto a compact star
We compute the energy gained by accretion onto a compact star due to matter falling from the inner edge of a thin accretion disk. We employ a controlled slow-rotation expansion based on the Hartle-Thorne metric and find that in neutron stars in general only leading order rotational corrections to the metric, describing frame dragging, are sizable. However, for accretion onto a magnetized neutron star, where the disk roughly extends to the co-rotation radius, even frame dragging effects are minor. Based on general conservation laws we then derive simple, equation-of-state-independent expressions for the spin-up and heating energy, consistently including the relevant rotational, relativistic and nuclear effects, and find that in observed accreting millisecond sources the results can strongly deviate from presently employed estimates.
Read the full abstract
Quantum Physics
EFI Pairs Without One-Way Puzzles: Oracle Separations from Communication Complexity
EFI pairs (Brakerski, Canetti, and Qian, ITCS 2023) and one-way puzzles (Khurana and Tomer, STOC 2024) are the leading candidates for the minimal assumption of quantum cryptography. The first are efficiently preparable quantum states, statistically far yet computationally indistinguishable; the second are classical puzzles, easy to sample and hard to solve. One-way puzzles imply EFI pairs, and whether the converse holds is open. We construct a single classical oracle relative to which one-way puzzles do not exist, even with an unbounded verifier, while an EFI pair survives every distinguisher that queries the oracle classically throughout and holds advice about it, making its one superposition query at the end. The oracle answers every question about the output probabilities of quantum samplers, which removes the puzzles, and hides a Haar-random half-dimensional subspace. To prove security we reduce it to communication complexity. An adversary whose knowledge of the subspace arrives as classical query answers can be simulated inside a two-party protocol against the party holding it, so it does no better than the best classical protocol for Vector-in-Subspace (Klartag and Regev, STOC 2011), whatever the oracle computes. That argument does not cover the superposition query, which we bound instead using tools from random matrix theory. The same attack gives a classical simulation of any quantum party in a classical-message protocol with no entanglement shared in advance, so relative to the oracle there is no proof of quantumness either. Quantum polynomial time therefore offers no advantage on any task with classical inputs and outputs, while the two quantum states stay indistinguishable. We state conjectures on removing the restriction on superposition queries.
Read the full abstract
Artificial Intelligence
MindTopo: Can Foundation Models Reason in Topological Space?
Spatial reasoning depends not only on metric properties such as distance, angle, and shape, but also on topological relations that remain invariant under continuous deformation. Cognitive science identifies these relations as foundational to spatial understanding, yet foundation-model evaluations largely focus on metric or viewpoint-dependent relations. We introduce MindTopo, a benchmark of topological intuition across five properties grounded in cognitive science and formal topology: continuity, separation, order, enclosure, and knots. MindTopo evaluates each property at two cognitive levels. Reasoning asks a model to identify topological relations or infer how they change. Planning instantiates a foundation model as a closed-loop agent whose policy selects environment actions. MindTopo contains 11,030 instances across 13 procedurally generated task types with controllable difficulty. We benchmark 14 MLLMs and study agent configurations augmented with image and video generation, including 3 video generative models in planning settings. Every MLLM performs better on reasoning than on planning, and the best-performing model remains far below observed human performance. On Qwen3-VL-2B-Instruct, supervised fine-tuning and reinforcement learning improve reasoning more than planning. Generated observations retain local cues and reach plausible endpoints, but audited rollouts do not reliably follow environment dynamics or preserve topology across transitions. Our website is at https://mind-topo.github.io/